[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <7027fa91-296f-01cb-6c5d-d25c81c835d5@redhat.com>
Date: Mon, 11 Oct 2021 10:34:39 +0200
From: David Hildenbrand <david@...hat.com>
To: Faiyaz Mohammed <quic_faiyazm@...eaurora.org>,
akpm@...ux-foundation.org, linux-mm@...ck.org,
linux-kernel@...r.kernel.org
Cc: guptap@...eaurora.org, Faiyaz Mohammed <faiyazm@...eaurora.org>
Subject: Re: [PATCH v2] mm: page_alloc: Add debug log in free_reserved_area
for static memory
On 11.10.21 09:11, Faiyaz Mohammed wrote:
> From: Faiyaz Mohammed <faiyazm@...eaurora.org>
>
> For INITRD and initmem memory is reserved through "memblock_reserve"
> during boot up but it is free via "free_reserved_area" instead
> of "memblock_free".
> For example:
> [ 0.294848] Freeing initrd memory: 12K.
> [ 0.696688] Freeing unused kernel memory: 4096K.
>
> To get the start and end address of the above freed memory and to account
> proper memblock added pr_debug log in "free_reserved_area".
> After adding log:
> [ 0.294837] 0x00000083600000-0x00000083603000 free_initrd_mem+0x20/0x28
> [ 0.294848] Freeing initrd memory: 12K.
> [ 0.695246] 0x00000081600000-0x00000081a00000 free_initmem+0x70/0xc8
> [ 0.696688] Freeing unused kernel memory: 4096K.
>
> Signed-off-by: Faiyaz Mohammed <faiyazm@...eaurora.org>
> ---
> changes in v2:
> - To avoid confusion, remove the memblock_dbg print and drop the
> memblock_free string, now using pr_debug to print the address ranges.
>
> mm/page_alloc.c | 5 +++++
> 1 file changed, 5 insertions(+)
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 668edb1..395df3f 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -8153,6 +8153,11 @@ unsigned long free_reserved_area(void *start, void *end, int poison, const char
> if (pages && s)
> pr_info("Freeing %s memory: %ldK\n", s, K(pages));
>
> +#ifdef CONFIG_HAVE_MEMBLOCK
> + pr_debug("%#016llx-%#016llx %pS\n",
> + __pa(start), __pa(end), (void *)_RET_IP_);
> +#endif
> +
Are we missing parentheses that the code indentation implies?
Further, I think we want to use "%pa" instead of "%llx" and
eventually drop the CONFIG_HAVE_MEMBLOCK dependency. Maybe
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index b37435c274cf..e7946e5ca094 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -8097,6 +8097,8 @@ EXPORT_SYMBOL(adjust_managed_page_count);
unsigned long free_reserved_area(void *start, void *end, int poison, const char *s)
{
+ const phys_addr_t pstart = __pa(start);
+ const phys_addr_t pend = __pa(end);
void *pos;
unsigned long pages = 0;
@@ -8125,9 +8127,11 @@ unsigned long free_reserved_area(void *start, void *end, int poison, const char
free_reserved_page(page);
}
- if (pages && s)
+ if (pages && s) {
pr_info("Freeing %s memory: %ldK\n",
s, pages << (PAGE_SHIFT - 10));
+ pr_debug("[%pa-%pa] %pS\n", &pstart, &pend, (void *)_RET_IP_);
+ }
return pages;
}
--
Thanks,
David / dhildenb
Powered by blists - more mailing lists